ABSTRACT The long-term goal is to develop ligand discovery methods, applying these to questions of active biological interest. We have four foci: i. development of new computational docking methods, testing these in simple experimental model systems (supported by GM59957). ii. Application of these methods to G Protein Coupled Receptors (GPCRs), which are intensely studied for the biology they confer, and wonderful templates for large- scale docking (U19 GM106990). iii. Investigation of the mechanism and impact of colloidal aggregation in drug discovery. These small molecule colloids, are the greatest source of artifacts in early discovery, and affect molecules throughout the drug development pipeline (GM71630). iv. Whereas our first three foci adopt a target-based view of ligand discovery, our fourth area returns to classical pharmacology, adopting a ligand- based chemoinformatic strategy that seeks not to discover new ligands for established targets, but rather for established drugs and reagents attempts to predict targets. This project has been the venue for the public access tools for chemoinformatics, databases, and docking (GM71896). Here we extend these projects. i. New docking methods are developed, including treating ordered waters, covalent recognition, and ligand internal energy strain. These are tested experimentally?by calorimetry and crystallography?in the model cavity sites. ii. Working with the Kobilka and Roth labs, we seek novel chemotypes for GPCRs including the - opioid receptor, the muscarinic M2 and M3 receptors, and the ?2-adrenergic receptors; a particular focus are allosteric ligands. iii. We deepen our investigation of the physical mechanism and impact colloidal aggregates, focusing on their persistence among late stage clinical candidates, their structure and mechanism, and antibody conjugates that specifically deliver colloidal drug payloads to cells. iv. Tool and database development remains a key focus. A new direction for the chemoinformatics is a comprehensive and systematic comparison of targets organized by sequence or by bioinformatics similarity, to the same targets organized by the similarity of their ligands. Preliminary results suggest target pairs organized by sequence, co- expression or protein-protein interactions are orthogonal to pairs related by similar ligands. A physical basis is explored.